Anthropic once warned the world about the dangers of powerful AI. Now that same technology drives its explosive growth. The San Francisco company just closed a massive funding round. Its valuation hit $965 billion. That figure tops OpenAI. Revenue run rate crossed $47 billion. All this happened while the firm kept tight controls on its most potent systems.
Months ago experts called one particular model too risky for wide release. Anthropic Turns Its Scariest AI Into A Money Printer reported how that model, Mythos, now anchors new security partnerships with 150 companies across 15 countries. Critical infrastructure operators in power, water, healthcare, communications and hardware gain access under strict rules. The goal? Protect vital software from the very threats advanced AI could unleash.
But don’t mistake caution for hesitation. Demand for Claude models surged. Enterprises snapped them up for coding and agentic tasks. Eight of the Fortune 10 count among customers. More than 1,000 enterprises pay over $1 million annually. Growth like this forces hard choices on compute allocation. Some users face limits. The company strains to keep pace.
Revenue tells the story best. Annual recurring revenue tripled from $9 billion earlier in the year to $47 billion by late May. The Wall Street Journal detailed expectations of a 130% surge to $10.9 billion in the June quarter. That milestone would deliver the firm’s first operating profit. Skeptics who doubted AI economics now confront numbers that defy them. Short bursts of acceleration. Then sustained scale. This pattern repeats.
Funding followed performance. Anthropic raised $65 billion in its Series H. The New York Times put pre-money valuation near $900 billion. Post-money reached $965 billion. Lead investors included Altimeter Capital, Dragoneer, Greenoaks Capital and Sequoia Capital. Co-leads featured Capital Group, Coatue, D1 Capital Partners, GIC, ICONIQ and XN. Hyperscalers contributed $15 billion, including $5 billion from Amazon. The official announcement on Anthropic’s site lists dozens more participants from Blackstone to Temasek. Funds will expand compute, push safety research and scale enterprise offerings.
Krishna Rao, Anthropic’s chief financial officer, captured the moment. “This funding will help us serve the historic demand we are experiencing, stay at the research frontier and bring Claude to more of the places where work happens.” Brad Gerstner of Altimeter added perspective. “This momentum positions Anthropic to lead the next phase of AI innovation and capture the enormous opportunity ahead.” Marc Stad at Dragoneer noted the progress feels breathtaking. “We believe that we are still in the earliest days of both the development and commercialization of this technology.”
Safety remains core. The company released Mythos only to select partners after rigorous screening that included the federal government. It refused Pentagon demands to loosen guardrails for “all lawful uses.” Officials sought flexibility for potential autonomous weapons or mass surveillance. Anthropic said no. A $200 million contract collapsed. The White House once labeled the firm a security risk. Ties later improved. Yet the episode highlights real friction. Commercial success collides with principles. Executives won’t bend on certain red lines.
That stance carries costs. It also builds trust. Enterprises wary of unchecked models turn to Claude. They value the firm’s responsible scaling policy and interpretability work. Recent updates to that policy refined thresholds for biological and chemical risks. The firm launched a fellows program to attract talent on scalable oversight, adversarial robustness and mechanistic interpretability. These moves signal long-term commitment. They don’t slow revenue.
New models keep coming. Claude Opus 4.8 arrived with the funding news. It excels at code generation. Performance gains outpace predecessors. Customers deploy it for complex agentic workflows. The New York Times noted how such advances help Anthropic pull ahead in enterprise adoption. Coding remains the killer application today. Tomorrow could shift toward broader autonomy. The firm prepares either way.
Profitability changes everything. First operating profit in the current quarter would silence doubters. Cash flow funds more research. It reduces reliance on ever-larger capital raises. IPO talks circulate. The company filed confidential S-1 papers. Timing remains fluid. Strong numbers give leverage with public markets. Investors who poured in at these valuations expect returns. So far the bet looks sound.
And risks persist. Dario Amodei has warned about concentrated power and wealth from AI. He speaks of trillionaires and political backlash if gains concentrate too fast. The firm donates to groups pushing regulation. It spent millions lobbying on safety standards. These actions match its founding ethos. Former OpenAI researchers started Anthropic precisely to prioritize caution. That origin story now underpins a near-trillion-dollar business.
Mythos captures the paradox. Once deemed one of the most dangerous technologies, it now secures software against future AI attacks. Project Glasswing coordinates controlled deployment. Early partners test defensive applications. Success here could spawn defense contracts and broader adoption. Failure invites criticism that the firm hyped dangers to sell solutions. Balance matters. So does transparency.
Competition intensifies. OpenAI pushes its own models and partnerships. xAI grows fast on different bets. Yet Anthropic’s mix of frontier performance, safety focus and enterprise traction sets it apart. Revenue run rate already exceeds many legacy software giants. Growth rate looks unsustainable until it sustains again. Quarter after quarter the pattern holds.
Compute hunger defines the next phase. Hyperscaler commitments help. Deals with Amazon, Google and Microsoft provide capacity. Still, demand outstrips supply. The firm limits access for some users. It prioritizes high-value enterprise workloads. This selectivity reinforces pricing power. Customers pay premiums for reliability and controls.
Executives sound measured. They celebrate progress without overclaiming. Quotes from the funding round emphasize early days. That humility contrasts with hype elsewhere in AI. It resonates with buyers who remember past technology cycles. They want results, not slogans. Anthropic delivers both capability and restraint.
The road ahead includes regulatory scrutiny. Political donations and lobbying draw attention. Relations with Washington evolve. Any IPO will face questions on governance, safety testing and military applications. The firm must prove its principles scale with its balance sheet.
For now momentum favors Anthropic. Scary models generate real revenue. Cautious policies attract serious customers. Massive capital flows in. Profitability appears within reach. The combination surprises even some insiders. It shouldn’t. The market rewards those who solve hard problems while addressing harder ones. Anthropic bet that approach would pay. Evidence mounts that it does.
WebProNews is an iEntry Publication